1. technology and computing

www.advmat.de
www.MaterialsViews.com
COMMUNICATION
Planar CH3NH3PbI3 Perovskite Solar Cells with Constant
17.2% Average Power Conversion Efficiency Irrespective of
the Scan Rate
Jin Hyuck Heo, Dae Ho Song, Hye Ji Han, Seong Yeon Kim, Jun Ho Kim, Dasom Kim,
Hee Won Shin, Tae Kyu Ahn, Christoph Wolf, Tae-Woo Lee, and Sang Hyuk Im*
Since Kojima et al.[1] reported CH3NH3PbX3 (MAPbX3, X = Br,
I) perovskite-sensitized solar cells, the perovskite hybrid solar
cells have been of great interest because of their unique properties such as long charge carrier’s diffusion length, strong
light absorptivity, ambipolar charge transport, high open circuit voltage, and solution processibility. The perovskite hybrid
solar cells can be roughly classified to mesoscopic,[2] mesosuperstructured,[3] and planar type.[4] The first uses the mesoscopic TiO2 electron conductor which can tune the electron flux
coming from the hybrid perovskite light absorber. Generally,
pure or alloyed perovskite hybrid solar cells adopt this device
architecture because the electrons generated in perovskite are
injected into mesoscopic TiO2. Meanwhile, the second uses the
mesoscopic insulator such as Al2O3 and as a result, the electrons are transported to electrode through the perovskite itself
because the electron injection from perovskite into Al2O3 is
blocked. The mixed halide perovskites such as MAPbI3-xClx
(x = 0.01–0.03) are usually chosen for this device structure. The
last utilizes the dense thin electron conducting layer such as
TiO2, ZnO, PCBM ([6,6]-phenyl-C61-butyric acid methyl ester)
instead of mesoscopic TiO2, of which the electrons generated
in planar perovskite thin film are injected into the electron conductor. All of pure, alloyed, and mixed halide perovskites can be
used for this planar type of perovskite hybrid solar cell.
Although the record efficiency of perovskite hybrid solar
cells reached up to 20.1%,[5] it was known that the perovskite
hybrid solar cells often exhibited significant hysteresis of J–V
J. H. Heo, D. H. Song, H. J. Han, Prof. S. H. Im
Functional Crystallization Center (FCC)
Department of Chemical Engineering
Kyung Hee University
1732 Deogyeong-daero, Giheung-gu, Yongin-si,
Gyeonggi-do 446-701, Republic of Korea
E-mail: [email protected]
S. Y. Kim, Prof. J. H. Kim
Department of Physics
Incheon National University
119 Academy-ro, Yeonsu-gu, Incheon 406-772, Republic of Korea
D. Kim, Dr. H. W. Shin, Prof. T. K. Ahn
Department of Energy Science
Sungkyunkwan University
Seobu-ro 2066, Jangan-gu, Suwon 440-746, Republic of Korea
C. Wolf, Prof. T.-W. Lee
Department of Materials Science and Engineering
Pohang University of Science and Technology (POSTECH)
77 Cheongam-Ro, Nam-Gu, Pohang,
Gyungbuk 790-784, Republic of Korea
DOI: 10.1002/adma.201500048
Adv. Mater. 2015,
DOI: 10.1002/adma.201500048
(photocurrent density–voltage) curves with respect to the scan
direction due to the charge accumulation or dielectric polarization by ferroelectric properties of perovskite materials.[6,7] It is
quite important issue to reduce the J–V hysteresis with the scan
direction in order to determine the real device efficiency. Jeon
et al.[2h] reported that the 200–300-nm thick mesoscopic TiO2
layer is necessary to reduce J–V hysteresis with respect to scan
direction because the electron–hole diffusion lengths are corresponded to 100–200 nm and the mesoscopic TiO2 electrode can
adjust the flux of electrons (Je) due to the its controlled interface
area.[2h] In solar cell, the highest device efficiency is obtained
when the electron flux (Je) and the hole flux (Jh) is balanced,
of which the flux stands for the number of charge carriers per
unit area and time. Otherwise, the unbalanced charge carriers
should be recombined by the direct recombination of electrons
and holes or the retarded recombination of charge carriers due
to the trapping at trap sites. If the material is 100% pure without
any traps, the unbalanced charge carriers should be recombined by direct recombination such as radiative or nonradiative
decay. Therefore, the traps are closely related to the response
time of solar cells and the mesoscopic cells might have advantage to balance the Je and Jh due to the controllable interface
area than the planar cells. This means that the unbalance of Je
and Jh leads charge accumulation to traps and the accumulated
charge carriers at traps make J–V hysteresis with respect to the
scan direction. In mesoscopic TiO2/MAPbI3 perovskite hybrid
solar cells, the major surface traps might be located at mesoscopic TiO2/MAPbI3 interface due to many junctions between
TiO2 nanocrystals. Therefore, we should consider both interface area and number of surface traps to minimize J–V hysteresis. So many junctions by thick mesoscopic TiO2 is good
for increasing Je but is bad for reducing J–V hysteresis due to
increased surface traps.
On the other hand, although the planar perovskite hybrid
solar cells have the simplest device architecture like bilayer
organic solar cells and the low temperature solution process is
possible due to the absent of mesoscopic TiO2 electrode which
generally requires hot processing temperature, it often suffers
from the severe J–V hysteresis with respect to the scan direction. For instance, recently Zhou et al.[4i] achieved 19.3% planar
MAPbI3-xClx mixed halide perovskite hybrid solar cells via spincoating method under controlled humid atmosphere. However, it revealed significant J–V hysteresis with respect to the
forward and reverse scan direction. Accordingly, it is difficult
to exactly determine the power conversion efficiency of planar
perovskite hybrid solar cells. It is also known that the formation of perovskite hybrid thin film with full surface coverage on
© 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
1
www.advmat.de
COMMUNICATION
www.MaterialsViews.com
the traps whereas the planar TiO2 electron
conductor cannot. The balanced electron and
hole flux will not create charge accumulation
within MAPbI3 perovskite layer if the charge
transport is not significantly hindered in the
electron conductor or the hole conductor.
Accordingly, it will be crucial to form pure
MAPbI3 perovskite layer in order to reduce
the J–V hysteresis with respect to scan direction in the planar MAPbI3 perovskite hybrid
solar cells.
As a model system for the formation of
pinhole-free MAPbI3 perovskite thin film
on TiO2/FTO substrate via single step spincoating process, we used 40 wt% MAPbI3/
DMSO (1 mL) solution and 40 wt% MAPbI3/
DMF (1 mL) solution with 100 µL HI addiFigure 1. a) Schematic illustration of planar MAPbI3 perovskite hybrid solar cell, b) its band tive solution, respectively, because the DMSO
solvent and DMF solvent with HI additive
energy diagram, and c) the SEM cross-sectional image of representative device.
have better solubility to MAPbI3 perovskite
than the pure DMF solvent thereby facilitating the formation
the electron conductor/FTO (F doped SnO2) substrate is tricky.
of pinhole-free thin film.[2h,4d] During a spin-coating process, a
Here, we report on the straightforward method for the formation of highly pure perovskite thin film with full surface covMAPbI3 perovskite solution are spread on the TiO2/FTO suberage thereby not only reducing significantly the J–V hysteresis
strate by centrifugal force and the MAPbI3 perovskite is crystalof perovskite hybrid solar cells with respect to the scan direclized on the substrate from the periphery to the center of the
tion but also demonstrating the 17.2% planar CH3NH3PbI3
substrate by further evaporation of solvent.[4d] At the moment,
perovskite solar cells with constant average power conversion
more concentrated MAPbI3 perovskite solution in DMSO and
efficiency irrespective with the scan rate.
DMF with HI additive can form more dense MAPbI3 thin film
Figure 1a depicts the device structure comprised FTO/
on the whole surface of TiO2/FTO substrate due to the increased
dense TiO2 electron conductor (≈50 nm)/MAPbI3 perovskite
solubility. Figure 2a,b shows the SEM surface images of MAPbI3
(≈300 nm)/PTAA (≈40 nm)/Au. In a typical process for the
perovskite thin film deposited on TiO2/FTO substrate via simple
fabrication of the perovskite hybrid device, a dense TiO2 elecone-step spin-coating process with MAPbI3/DMSO (1 mL) solutron conductor of ≈50 nm in thickness was deposited on a FTO
tion and MAPbI3/DMF (1 mL) solution with 100 µL HI addiglass by the spray pyrolysis deposition method. The pinholetive solution. The SEM images clearly confirm the formation of
free MAPbI3 thin films were deposited on the TiO2/FTO subdense MAPbI3 thin film without pinholes. To compare the crysstrate by spin-coating of 40 wt% MAPbI3/DMF (N,N-dimethyltalline morphology of whole surface of DMSO- and HI-MAPbI3
formamide: 1 mL) with 100 µL HI or 40 wt% MAPbI3/DMSO
film, we checked SEM surface images of them in detail as
shown in Figure S1, Supporting Information. We found that the
(dimethyl sulfoxide: 1 mL) solutions at 3000 rpm for 200 s.
DMSO-MAPbI3 film has very large crystalline domains, whereas
Then, the PTAA solution was spin-coated on MAPbI3/TiO2/
FTO substrate at 3000 rpm for 30 s. Finally, the Au counter
the HI-MAPbI3 film has not any crystalline domains. Their corelectrode was deposited by thermal evaporation. The representresponding SEM cross-sectional images in Figure 2c,d indicate
ative SEM cross-sectional image of planar MAPbI3 perovskite
that the thickness of both MAPbI3 layers is ≈300 nm. To check
hybrid solar cell was shown in Figure 1c.
the purity of MAPbI3 perovskite layer formed by MAPbI3/DMSO
Upon illumination of light, the MAPbI3 perovskite generates
solution (DMSO-MAPbI3) and MAPbI3/DMF solution with HI
electron–hole pairs and the electrons (holes) are then transadditive solution (HI-MAPbI3), the XRD patterns were shown in
ported into TiO2 electron conductor (PTAA hole conductor)
Figure 2e. The XRD patterns clearly indicate that the MAPbI3
as shown in Figure 1b. If the MAPbI3 perovskite layer has
perovskite film formed by MAPbI3/DMSO solution has some
some defects, the photocurrent response will be delayed by
impurities such as PbI2 whereas no impurities are detectable in
the internal trap states because the traps act as internal capaciMAPbI3 film formed by MAPbI3/DMF solution with HI additive.
tance due to the trap filling and trapping/detrapping process.[8]
Although Jeon et al.[2h] reported that the DMSO additive does not
Therefore, the impurities in MAPbI3 perovskite layer will retard
deteriorate the purity of MAPbI3 by nonsolvent dripping process,
the charge injection from MAPbI3 into TiO2. In addition, the
tiny PbI2 impurity peaks always existed in the closer look of XRD
capacitance element in the impure MAPbI3 will make severe
patterns. Similarly, the MAPbI3 crystals formed by in MAPbI3/
J–V hysteresis with respect to scan direction. Especially, the
DMF solution (DMF-MAPbI3) revealed also small amount of
planar MAPbI3 perovskite hybrid solar cells are expected to
PbI2 impurity peaks in Figure 2e. The formation of impurities
reveal more severe J–V hysteresis than the mesoscopic MAPbI3
in MAPbI3 film formed by the conventional MAPbI3/DMF and
perovskite solar cells[9,10] because the mesoscopic TiO2 electron
MAPbI3/DMSO solutions might be associated with the fact
conductor can compensate the electron flux due to the large
that practically, it is very difficult to exactly mix equimolar consurface area even though the electron injection is retarded by
centration of MAI and PbI2 or the MAI can be decomposed or
2
wileyonlinelibrary.com
© 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2015,
DOI: 10.1002/adma.201500048
www.advmat.de
www.MaterialsViews.com
COMMUNICATION
Figure 3a is UV–vis absorption spectra
of FTO/TiO2/MAPbI3 films prepared from
DMSO-MAPbI3 and HI-MAPbI3 solutions
indicating that both MAPbI3 films have
similar light harvesting efficiency due to
the same thickness of MAPbI3 layer (see
Figure 2c,d). Figure 3b exhibits the representative J–V curves of planar MAPbI3
perovskite hybrid solar cells prepared by
DMSO-MAPbI3 and HI-MAPbI3 solutions
with respect to scan direction under 200 ms
of delay time and 10 mV per point (voltage
scan step) condition. As expected from the
XRD analysis, the planar MAPbI3 perovskite hybrid solar cells prepared by DMSOMAPbI3 had more serious J–V hysteresis
with respect to the scan direction than the
cell by HI-MAPbI3 solutions. The photovoltaic properties of DMSO-MAPbI3 (DMSO)
and HI-MAPbI3 (HI) planar perovskite
hybrid solar cells were listed in Table 1.
The DMSO-MAPbI3 device exhibited that
1.06 V of Voc, 18.7 mA cm−2 of Jsc, 59% of
F.F., and 11.7% of power conversion efficiency (η) at forward scan direction at 1 sun
condition and 1.08 V of Voc, 18.9 mA cm−2
of Jsc, 70% of F.F., and 14.3% of η at backward scan direction thereby yielding to 13%
of average power conversion efficiency (ηavg).
On the other hand, the HI-MAPbI3 device
exhibited 1.1 V of Voc, 20.4 mA cm−2 of Jsc,
75% of F.F., and 16.8% of η at forward scan
direction and 1.1 V of Voc, 20.5 mA cm−2 of
Jsc, 78% of F.F., and 17.6% of η at backward
scan direction, thereby yielding to 17.2%
of ηavg. These results indicate that the pure
MAPbI3 perovskite can reduce the J–V hysteresis with respect to scan direction and
Figure 2. SEM surface a,b) and cross-sectional c,d) images of MAPbI3 perovskite thin films increase the average power conversion effiprepared by one-step spin-coating process using MAPbI3/DMSO solution (DMSO-MAPbI3: (a) ciency as well. The improved efficiency
and (c)) and MAPbI3/DMF solution with HI additive solution (HI-MAPbI3: (b) and (d)); and
might be attributed to the enhanced charge
e) XRD patterns of MAPbI3 perovskite thin films (DMF-MAPbI3 = MAPbI3/DMF solution
injection and/or charge collection efficiency
without HI additive solution).
because the HI-MAPbI3 hybrid solar cells
exhibited better EQE (external quantum efficiency) spectrum than the DMSO-MAPbI3 cells as shown in
sublimated during the formation of MAPbI3 crystal film. Therefore, the mesoscopic MAPbI3 perovskite solar cells exhibited
Figure 3a. The photovoltaic parameters of DMSO-MAPbI3 and
small hysteresis of short circuit current density (Jsc), open cirHI-MAPbI3 planar hybrid solar cells with respect to scan direccuit voltage (Voc), fill factor (F.F.), and efficiency even though the
tion under different delay time conditions were summarized in
Table S1, Supporting Information. The power conversion effithickness of mesoscopic TiO2 is optimized.[2h] On the other hand,
ciencies of DMSO-MAPbI3 and HI-MAPbI3 planar hybrid solar
the added HI solution can reduce the formation of PbI2 impurities by decomposition of MAI from MAPbI3 because the HI
cells with respect to scan direction at different delay time condition were plotted in Figure 3c. This clearly shows that the J–V
solution can recover the methylamine (decomposed product of
hysteresis of HI-MAPbI3 planar hybrid solar cells with respect
MAI) into MAI thereby significantly suppressing the decomposition reaction of MAPbI3 perovskite. Accordingly, it is expected
to scan direction is significantly reduced. The HI-MAPbI3
that the planar MAPbI3 perovskite hybrid solar cells with high
planar hybrid solar cells had the coincidence of power conversion efficiencies measured by forward and backward scan conpurity will exhibit smaller J–V hysteresis with respect to scan
dition over 500 ms of delay time whereas the DMSO-MAPbI3
direction and scan rate. The effect of PbI2 impurity on the J–V
hysteresis with respect to the scan direction is demonstrated in
cells had over 1000 ms of delay time. It should be noted that
Figure S2, Supporting Information.
the HI-MAPbI3 planar hybrid solar cells exhibited constant ηavg
Adv. Mater. 2015,
DOI: 10.1002/adma.201500048
© 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
3
www.advmat.de
COMMUNICATION
www.MaterialsViews.com
Figure 3. a) Absorption and EQE (external quantum efficiency) spectra of DMSO-MAPbI3 and HI-MAPbI3 film and device, respectively and b) J–V
curves, c) forward and backward efficiency with delay time, and d) efficiency deviation of DMSO-MAPbI3 and HI-MAPbI3 perovskite hybrid solar cells.
(DMSO = DMSO-MAPbI3 device, HI = HI-MAPbI3 device, forward = forward scan, backward = backward scan.)
irrespective to the delay time whereas the DMSO-MAPbI3 cells
showed slightly declining ηavg with the delay time. Figure 3d
is the ηavg deviation of different 30 DMSO-MAPbI3 and HIMAPbI3 hybrid solar cells, respectively, indicating that the ηavg
is significantly improved by introduction HI additive which
enables to form pure MAPbI3 perovskite thin film by simple
one-step spin-coating process. It is also noted that the existence of J–V hysteresis with respect to the scan direction and
scan rate do not make serious problem if we measure correctly
measure the efficiency of perovskite hybrid solar cells through
sufficiently increasing the delay time (>1000 ms) where the J–V
curves of forward and backward scan condition are coincided.
Hence, we need to carefully check the J–V curves in perovskite
hybrid solar cells to exactly determine the real efficiency. The
current density under applied optimal bias voltage at maximum
power point with light soaking time shown in Figure S3, Supporting Information, clearly confirms that the HI-MAPbI3
hybrid solar cell has a constant power conversion efficiency
under continuous light illumination of 1 sun, whereas the
Table 1. Photovoltaic properties of DMSO-MAPbI3 (DMSO) and HIMAPbI3 (HI) planar perovskite hybrid solar cells (100 mW cm−2 AM
1.5G, delay time = 200 ms, voltage scan step = 10 mV).
Device
DMSO
HI
4
Scan direction
Voc
[V]
Jsc
[mA cm−2]
F.F.
[%]
η
[%]
Avg. η
[%]
13.0
Forward
1.06
18.7
59
11.7
Backward
1.08
18.9
70
14.3
Forward
1.1
20.4
75
16.8
Backward
1.1
20.5
78
17.6
wileyonlinelibrary.com
17.2
DMSO-MAPbI3 solar cell’s efficiency is slightly declined with
light soaking time.
To understand the effect of HI additive on the device efficiency and J–V hysteresis due to the higher impurities which
might be associated with the traps, the trap density of completed solar cell was investigated by using admittance spectroscopy (AS). AS can probe trap energy level and trap density of
the absorber layer.[11] The trap energy level is derived by the
⎛ −E ⎞
relation, ω 0 = βT 2 exp ⎜ a ⎟ , where ωo is characteristic transi⎝ kBT ⎠
tion frequency, β is temperature independent parameter, Ea is
depth of trap energy level, kB is the Boltzmann constant, and T
is the temperature. Using this equation, Arrhenius plot of characteristic frequency is plotted, and Ea is the slope of the Arrhenius plot line. The trap density can be derived by the equation,
Vbi dC ω
, where C is the capacitance, ω is the applied
Nt = −
qw dω kBT
frequency, q is the elementary charge, kB is the Boltzmann constant, T is the temperature, Vbi is the built-in potential, and w
is the depletion width, respectively. The values of Vbi and w are
obtained from Mott–Schottky analysis.
Figure 4a–f shows the results of AS measurements.
DMSO solar cell shows trap energy level of 0.27 eV and trap
energy density of ≈1017 cm−3, whereas HI solar cell exhibits
trap energy level of 0.28 eV and trap energy density of
≈1016 cm−3. The trap energy of 0.27–0.28 eV is shallower than
that reported by Shao et al.[12] and deeper than the result by
Duan et al.[13] It should be noted that trap density of DMSO
cell is 10 times larger than that of HI cell. The higher defect
density induces larger electrical element of capacitance in the
© 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2015,
DOI: 10.1002/adma.201500048
www.advmat.de
www.MaterialsViews.com
COMMUNICATION
Figure 4. Temperature dependent capacitance spectra (a) for DMSO, d) for HI, Arrhenius plots of characteristic frequencies (b) for DMSO, e) for
HI, and trap density profiles (c) for DMSO, f) for HI of DMSO (a–c) and HI (d–f) cells, respectively; g) time-resolved photoluminescent (PL) decay
curves of DMSO and HI; and h,i) charge carrier’s diffusion coefficient from (h) IMPS and decay time from (i) IMVS of DMSO and HI device. (DMSO
= DMSO-MAPbI3 planar hybrid solar cell, HI = HI-MAPbI3 planar hybrid solar cell.)
complete solar cell, which may result in hysteresis of solar cell.
It should be noted that the capacitance elements are attributed to the impurities in MAPbI3 perovskite thin film within
itself because the charges can be accumulated at the trap
sites. Therefore, the traps are gradually filled at forward scan
condition, whereas the traps are fully filled at backward scan
condition. Under fast scan rate (short delay time) condition
of J–V curves, the photocurrent is measured before reaching
the saturation value so that the Jsc is increased with the delay
time at forward scan condition, whereas the Jsc is decreased
with the delay time due to the recombination through the trap
sites at backward scan condition. As a result, the η of DMSOMAPbI3 device reveals significant hysteresis with respect to
the scan direction, whereas the Jsc of HI-MAPbI3 device is less
sensitive with the scan rate than the DMSO-MAPbI3 device.
Consequently, the HI-MAPbI3 perovskite hybrid solar cells
exhibited constant ηavg irrespective to the scan rate meanwhile
the ηavg of DMSO-MAPbI3 device was slightly decreasing
with the delay time. In addition, we need to consider the
change of ferroelectric properties with different perovskite
Adv. Mater. 2015,
DOI: 10.1002/adma.201500048
thin films of DMSO-MAPI3 and HI-MAPbI3. If the change of
dielectric polarization by ferroelectric properties is dominant
factor to make J–V hysteresis, the J–V hysteresis were greatly
dependent on the perovskite crystalline domain size but here
the morphologies of perovskite thin films were similar so that
the traps might be the major factor to make J–V hysteresis
(see the more detailed explanation for AS analysis in the Supporting Information).
To compare the charge carrier dynamics (charge transfer
or charge injection/separation behavior), we measured the
time-resolved photoluminescent (PL) decay curves of DMSOMAPbI3 and HI-MAPbI3 planar hybrid solar cells as shown
in Figure 4g.[14] The shortened PL decay in HI-MAPbI3
solar cells confirmed the faster transfer of the charge carriers into TiO2 electron conductor and PTAA hole conductor
than that in the DMSO-MAPbI3 solar cells, which means
that HI-MAPbI3 perovskite hybrid solar cells can inject the
charge carriers from MAPbI3 peorvskite into TiO2 electron
conductor more efficiently. From the convolution of the timeresolved PL decays with a biexponential function, the average
© 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
5
www.advmat.de
www.MaterialsViews.com
COMMUNICATION
PL lifetime (τavg.) of HI-MAPbI3 (2.058 ns) is shorter than
that of DMSO-MAPbI3 (2.278 ns), as listed in Table S2, Supporting Information.
To check what the purity of MAPbI3 perovskite can affect the
device operation, we measured diffusion coefficient (Dn) and
charge carrier’s life time (τn) from IMPS (intensity modulated
photocurrent spectroscopy) and IMVS (intensity modulated
photovoltage spectroscopy) analysis as shown in Figure 4h,i.
The original Nyquist plots of IMPS and IMVS were shown in
Figure S4, Supporting Information. From the plots of Dn and
τn with current density (J), we could recognize that the Dn
and τn of HI-MAPbI3 planar hybrid solar cells are increased
to ≈1.3 and ≈1.6-fold, respectively, than the DMSO-MAPbI3
devices. Accordingly, the diffusion length (Ln = (Dn·τn)0.5) of
MAPbI3 planar hybrid solar cells was increased from ≈230 to
≈350 nm by using pure MAPbI3 perovskite (HI-MAPbI3). From
the analysis of transient PL decay, IMPS, and IMVS, we could
concluded that i) the EQE of HI-MAPbI3 planar hybrid solar
cell is improved by the enhanced charge injection/separation
efficiency and charge collection efficiency thereby improving
Jsc; ii) the F.F. and Voc of HI-MAPbI3 planar hybrid solar cell
is enhanced by the improved Dn and τn thereby reducing the
backward recombination of electrons and holes; and iii) the J–V
hysteresis of HI-MAPbI3 device with respect to the scan direction and rate is greatly reduced by the formation of highly pure
MAPbI3 perovskite by addition of HI additive solvent thereby
eliminating number of traps which may create the internal
capacitance elements within MAPbI3 perovskite layer.
In summary, the pinhole-free MAPbI3 perovskite hybrid
thin-films were formed by the simple one-step spin-coating
process by using MAPbI3/DMSO solution and MAPbI3/DMF
with HI additive solvent. The XRD analysis confirmed that the
pure MAPbI3 perovskite hybrid thin film could be formed by
the contribution of HI additive solvent because HI can prohibit the decomposition of MAPbI3 to MAI and PbI2 whereas
conventional MAPbI3 in DMF or DMSO solution made some
impurity such as PbI2. Through the formation of pure MAPbI3
perovskite hybrid thin films by introduction of HI additive
solvent, we could enhance 1.3-fold of the ηavg under illumination of 1 sun (100 mW cm−2 AM 1.5G) light and consequently
obtained 17.2% of ηavg irrespective with the scan direction
and rate because the HI-MAPbI3 planar perovskite hybrid
solar cells exhibited ≈1.4-fold longer diffusion length than the
DMSO-MAPbI3 planar solar cells owing to ≈1.3 and ≈1.6-fold
of the improved Dn and τn, respectively, than DMSO-MAPbI3
planar solar cells. In addition, the charge injection/separation efficiency and trap sites are also improved by the formation of pure MAPbI3 perovskite hybrid thin films. From the AS
results, we found that both DMSO- and HI-cells have similar
trap energy level of 0.27–0.28 eV but the HI cell has ≈10-fold
lower trap density than the DMSO cell. Therefore, the HI cell
reveals lower J–V hysteresis with respect to the scan direction
due to the smaller number of capacitance element. Therefore,
we believe that if we balance the electron (Je) and hole flux (Jh)
and eliminate the trap sites in MAPbI3 planar perovskite hybrid
solar cells, the highly efficient planar perovskite hybrid solar
cells without J–V hysteresis with respect to scan direction could
be fabricated.
6
wileyonlinelibrary.com
Experimental Section
Preparation of MAPbI3 Perovskite Solution: To synthesize CH3NH3I3,
50 mL hydriodic acid (48% in water, Aldrich) and 30 mL methylamine
(40% in methanol, Junsei Chemical Co. Ltd.) were reacted in a 250-mL
round-bottom flask at 0 °C for 2 h with vigorous stirring. The products
were recovered by evaporation and then were dissolved in ethanol,
recrystallized from diethyl ehter. Finally, the products were dried at the
room temperature in a vacuum oven for 24 h. To form the CH3NH3PbI3
thin film, two kinds of solutions such as 40 wt% of CH3NH3PbI3 in
dimethyl sulfoxide and N,N-dimethylformamide with added 100 µL
hydriiodic acid were prepared by reacting the synthesized CH3NH3I and
PbI2 (Aldrich) at 1:1 mole ratio at 60 °C for 30 min.
Device Fabrication: To fabricate the planar CH3NH3PbI3 perovskite
hybrid solar cell, dense TiO2 blocking layer (bl-TiO2) of ≈50 nm
thickness was deposited on cleaned F-doped SnO2 (FTO, Pilkington,
TEC8) glass substrate by spray pyrolysis deposition (SPD) method
with 20 × 10−3 M of titanium diisopropoxide bis(acetylacetonate)
(Aldrich) solution at 450 °C. To form the CH3NH3PbI3 layer on
bl-TiO2/FTO substrate, the 40 wt% CH3NH3PbI3 in dimethyl sulfoxide
and N,N-dimethylformamide solution with added 100 µL HI was then
coated by spin-coating and dried on a hot plate at 100 °C for 2 min.
In each solvent, the spin-coating conditions are at 2000 rpm for 60 s
then 4000 rpm for 120 s and at 3000 rpm for 200 s. Poly[bis(4-phenyl)(2,4,6-trimethylphenyl)amine] (PTAA: EM index) hole transporting
materials (HTM) were spin-coated on CH3NH3PbI3/bl-TiO2/FTO
substrate at 2000 rpm for 30 s by using HTM/toluene (15 mg/1 mL)
with 7.5 µL Li-bis(trifluoromethanesulfonyl) imide (Li-TFSI)/
acetonitrile (170 mg/1 mL) and 7.5 µL t-BP/acetonitrile
(1 mL/1 mL) additives and without additives. Finally, Au counter
electrode was deposited by thermal evaporation. The active area was
fixed to 0.16 cm2. All device fabrications were conducted below 50%
of relative humidity.
Device Characterization: The current density–voltage (J–V) curves were
measured by a solar simulator (Peccell, PEC-L01) with a potentiostat
(IVIUM, IviumStat) at 100 mA cm−2 illumination AM 1.5G and a
calibrated Si-reference cell certificated by JIS (Japanese Industrial
Standards). For the measurement hysteresis of J–V curves, the forward
and reverse scan rate was set to 200 ms/10 mV as a standard condition
and was varied from 100 ms/10 mV to 1000 ms/10 mV. The J–V
curves of all devices were measured by masking the active area with
metal mask of 0.096 cm2. The external quantum efficiency (EQE) was
measured by a power source (ABET 150W Xenon lamp, 13014) with a
monochromator (DONGWOO OPTORN Co., Ltd., MonoRa-500i) and
a potentiostat (IVIUM, IviumStat). AS measurements were carried out
by using precision LCR meter (Agilent, E4980A) under dark condition
at frequencies between 0.1 and 1000 kHz. Sample was located in homemade cryostat using liquid nitrogen, and the range of measurement
temperature was 150–300 K. As temperature was cooled down with
cooling rate of ≈2 K min−1, capacitance spectra were measured at every
10 K, and all measurements were done at corresponding temperature
with error bar of ±0.05 K or less.
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
Acknowledgements
This work was supported by the Global Frontier R&D Program of the
Center for Multiscale Energy System, Mid-career Research Program
(No. NRF-2013R1A2A2A01067999), and Basic Science Research
Program (No. 2014R1A5A1009799) through the National Research
© 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2015,
DOI: 10.1002/adma.201500048
www.advmat.de
www.MaterialsViews.com
Received: January 5, 2015
Revised: March 31, 2015
Published online:
[1] A. Kojima, K. Teshima, Y. Shirai, T. Miyasaka, J. Am. Soc. 2009, 131,
6050.
[2] a) H.-S. Kim, C. R. Lee, J. H. Im, K. B. Lee, T. Moehl, A. Marchioro,
S. J. Moon, R. Humphry-Baker, J. H. Yum, J. E. Moser, M. Grätzel,
N.-G. Park, Sci. Rep. 2012, 2, 591; b) J. H. Heo, S. H. Im, J. H. Noh,
T. N. Madal, C. S. Lim, J. A. Chang, Y. H. Lee, H.-J. Kim, A. Sarkar,
Md. K. Nazeeruddin, M. Grätzel, S. I. Seok, Nat. Photonics 2013,
7, 486; c) J. H. Noh, S. H. Im, J. H. Heo, T. N. Mandal, S. I. Seok,
Nano Lett. 2013, 13, 1764; d) J. H. Noh, N. J. Jeon, Y. C. Choi,
Md. K. Nazeeruddin, M. Grätzel, S. I. Seok, J. Mater. Chem. A
2013, 1, 11842; e) N. J. Jeon, J. Lee, J. H. Noh, M. K. Nazeeruddin,
M. Grätzel, S. I. Seok, J. Am. Chem. Soc. 2013, 135, 19087;
f) J. H. Heo, S. H. Im, Phys. Status Solidi 2014, 8, 816; g) J. Burschka,
N. Pellet, S. J. Moon, R. H. Baker, P. Gao, M. K. Nazeeruddin,
M. Grätzel, Nature 2013, 499, 316; h) N. J. Jeon, J. H. Noh,
Y. C. Kim, W. S. Yang, S. Ryu, S. I. Seok, Nat. Mater. 2014, 13, 897.
[3] a) M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami, H. J. Snaith,
Science 2012, 338, 643; b) J. M. Ball, M. M. Lee, A. Hey, H. J. Snaith,
Energy Environ. Sci. 2013, 6, 1739; c) D. Bi, S.-J. Moon, L. Häggman,
G. Boschloo, L. Yang, E. M. J. Johansson, Md. K. Nazeeruddin,
M. Grätzel, RSC Adv. 2013, 3, 18762.
[4] a) M. Liu, M. B. Johnston, H. J. Snaith, Nature 2013, 501, 395;
b) O. Malinkiewicz, A. Yella, Y. H. Lee, G. M. Espallargas,
M. Grätzel, M. K. Nazeeruddin, H. J. Bolink, Nat. Photonics 2014, 8,
Adv. Mater. 2015,
DOI: 10.1002/adma.201500048
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
COMMUNICATION
Foundation of Korea (NRF) funded by the Ministry of Science, ICT &
Future Planning.
128; c) D. Liu, T. L. Kelly, Nat. Photonics. 2014, 8, 133; d) J. H. Heo,
D. H. Song, S. H. Im, Adv. Mater. 2014, 26, 8179; e) Z. Xiao,
Q. Dong, C. Bi, Y. Shao, Y. Yuan, J. Huang, Adv. Mater. 2014, 26,
6503; f) J.-H. Im, I.-H. Jang, N. Pellet, M. Grätzel, N.-G. Park, Nat.
Nanotechnol. 2014, 9, 927; g) J. Seo, S. Park, Y. C. Kim, N. J. Jeon,
J. H. Noh, S. C. Yoon, S. I. Seok, Energy Environ. Sci. 2014, 7, 2642;
h) M. Xiao, F. Huang, W. Huang, Y. Dkhissi, Y. Zhu, J. Etheridge,
A. Gray-Weale, U. Bach, Y.-B. Cheng, L. Spiccia, Angew. Chem. Int.
Ed. 2014, 126, 10056; i) H. Zhou, Q. Chen, G. Li, S. Luo, T.-Z. Song,
H.-S. Duan, Z. Hong, J. You, Y. Liu, Y. Yang, Science 2014, 345, 542.
Research Cell Efficiency Records. NREL, http://www.nrel.gov/ncpv/
(accessed: December 2014).
H.-S. Kim, I. Mora-Sero, V. Gonzalez-Pedro, F. Fabregat-Santiago,
E. J. Juarez-Perez, N.-G. Park, J. Bisquert, Nat. Commun. 2013, 4,
2242.
J. M. Frost, K. T. Butler, F. Brivo, G. H. Hendon, M. v. Schilfgaarde,
A. Walsh, Nano Lett. 2014, 14, 2584.
K. Schwarzburg, F. Willig, Appl. Phys. Lett. 1991, 58,
2520.
A. Dualeh, T. Moehl, N. Tétreault, J. Teuscher, P. Gao,
Md. K. Nazeeruddin, M. Grätzel, ACS Nano 2014, 8, 362.
H. J. Snaith, A. Abate, J. M. Ball, G. E. Eprron, T. Leijtens,
N. K. Noel, S. D. Stranks, J. T.-W. Wang, K. Wojciechowski,
W. Zhang, J. Phys. Chem. Lett. 2014, 5, 1511.
T. Walter, R. Herberholz, C. Müller, H. Schock, J. Appl. Phys. 1996,
80, 4411.
Y. Shao, Z. Xiao, C. Bi, Y. Yuan, J. Huang, Nat. Commun., 2014, 5,
5784.
H.-S. Duan, H. Zhou, Q. Chen, P. Sun, S. Luo, T.-B. Song, B. Bob,
Y. Yang, Phys. Chem. Chem. Phys. 2014, 17, 112.
S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou,
M. J. P. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza, H. J. Snaith,
Science 2013, 342, 344.
© 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
7